SophNet | Cloud Compute Platform Officially Launches QwQ-32B Model with Inference Speed Reaching 75 TPS
On March 6, Alibaba's Tongyi Qianwen unveiled its latest inference model QwQ-32B, which achieves a quantum leap in mathematical, coding, and general capabilities. With overall performance rivaling DeepSeek-R1, it has reignited market enthusiasm for homegrown large language models.
SophNet platform delivered as expected, quickly launching the QwQ-32B large model. Leveraging the powerful computing capabilities of Sophgo's self-developed TPU, QwQ-32B achieves an inference speed of up to 75 Tokens/s—lightning-fast and more stable in performance.
SophNet QwQ-32B response effect: Lightning speed
Scan the QR code below to directly experience it on the SophNet Test Field：
Developers simply need to select the QwQ-32B model when creating a service to easily complete the API configuration.
In the future, SophNet will continue its mission to provide stable and high-speed inference services, continuously optimize the cloud computing experience, and enable enterprises and developers to effortlessly access global top AI models, accelerating the universal adoption of AI computing for humanity.
Platform Introduction
SophNet | Cloud Computing Platform (Website: sophnet.com) – Leveraging the powerful computing capabilities of Sophgo's self-developed TPU processors, we are committed to building a faster, more stable, and cost-effective one-stop model service platform.
With just a few lines of code, users can access the world's leading AI large models. They can create various workflow topologies tailored to their application needs, from simple API services to complex agent-based intelligent systems.
Currently, the platform integrates mainstream open-source large models such as DeepSeek R1 Full Version, QwQ-32B, DeepSeek V3, and Qwen2.5, and supports private model hosting. Through one-click deployment, enterprises and developers can enjoy powerful computing resources at ultra-low costs, achieving true model and computing power freedom, and accelerating the democratization of AI computing for all.
Contact Us
Website: https://sophnet.com
Business Cooperation: 010-57590723
Email: sophnet@sophgo.com
Enterprise WeChat:
